The Martha's Vineyard Child Health Survey is a 6 year study to determine the prevalence, duration and causes of psychiatric disorders in children during the first seven years of life. The project began in 1977 by examining the prevalence of behavior problems in 3 year old children. The frequency of psychiatric disorders was found to be 14%. In a subsequent analysis all of the children seclected to have behavior problems of moderate to severe degree could be assigned a diagnosis according to DSM-III criteria. Both social and temperamental factors have been associated with the presence of psychiatric disorders in this group of children. In 1981 a follow-up study of these 3 year olds was completed. The children were 6 or 7 years at the time of follow-up and in first or second grade. Particular emphasis was devoted to the development of social competence as well as to persitence of behavior problems. The major finding of this study is that social competence can best be predicted from demographic and intrafamilial variables, rather than from personality or behavioral variables. As aspect of the study which began in 1980 and continues to the present is the collection of life events and social support data in families of children 0-2 years who are participating in this phase of the project. When completed this propective, longitudinal study will use this data to predict the development of psychiatric disorders in very young children.